I was trying to turn some 1/2" 6061 aluminum rod down to 12mm on a 7x10 mini lathe, but it's not coming out the way I expected. i have about 1" of the rod in a 3 jaw chuck and 2" of the rod sticking out which I am turning down to 12mm.

I started turning and measuring the resulting diameter with a micrometer and noticed that the diameter closest to the chuck is narrower then the diameter of the end furthest from the chuck by about 0.001". I'm using the power feed and not touching the compound so the angle of the compound doesn't come into play.

Anyone have and ideas on what could be causing the problem and how to check for it? I don't have much experience with the lathe so I'm kind of at a loss.

Thanks!

Is the tool on center and what is your spindle speed?

Are you doing another/other cut at the same depth to remove the effect of deflection caused by the tool?

John

... 2" of the rod sticking out which I am turning down to 12mm.....the diameter closest to the chuck is narrower then the diameter of the end furthest from the chuck by about 0.001".....

0.001" on the diameter is 0.0005" on the radius. Over 2 inches of length on 12mm diameter 0.0005" deflection is to be expected. I would be surprised if you get much better than this even with the sharpest tool you can find. It is possible to reduce the deflection my going to a tool with a very small nose radius and taking the feed very slow to maintain a decent surface finish.

Is the tool on center and what is your spindle speed?

I'm turning at around 750 to 800 RPM, the speed knob is at the 12 o'clock position on the mini-lathe in low gear. I'm using a 5/16" tool bit which is on center.

Are you doing another/other cut at the same depth to remove the effect of deflection caused by the tool?

I don't think i did. I got kind of discouraged when I figured out the taper was happening and decided to call it quits for the day.

It is possible to reduce the deflection by going to a tool with a very small nose radius and taking the feed very slow to maintain a decent surface finish.

I will try a tool with a smaller nose radius. Good to know about the feed setting in the future. Right now I'm already running the slowest gear combination, but I'm trying to CNC this lathe so I can run slower in the future.

Can i compensate for the deflection by turning the piece twice without changing the cutting tool's position? I'm trying to cut a 2" in length 12mm shaft with -0" + 0.0005" tolerance. (I know, metric + english combination = bad). I'm not sure what the mini lathe is capable of with the stock 3 jaw chuck. Is this pushing it with the mini lathe?

Is the deflection caused by the tool bending or the workpiece bending? I'm trying to picture in my head what is going on.

.... Is this pushing it with the mini lathe?

Is the deflection caused by the tool bending or the workpiece bending? I'm trying to picture in my head what is going on.

It is not so much pushing the mini lathe but pushing the workpiece, quite literally :) .

The tool has to exert a force on the work in order to cut; it is impossible to do otherwise. If you want to cut something you have to exert a force. The force needed to cause the tool to cut is enough to deflect the small diameter, not very stiff metal, you are cutting especially over the length you are dealing with.

If it is imperative you have it parallel to within better tha .0005 on the radius turn it between centers. But then of course you may need to align your tailstock center to get this precision.

A perfectly legimiate alternative is to carefully polish the taper out with some very fine emery paper and delicate fingers.

I was trying to turn some 1/2" 6061 aluminum rod down to 12mm on a 7x10 mini lathe, but it's not coming out the way I expected. i have about 1" of the rod in a 3 jaw chuck and 2" of the rod sticking out which I am turning down to 12mm.

I started turning and measuring the resulting diameter with a micrometer and noticed that the diameter closest to the chuck is narrower then the diameter of the end furthest from the chuck by about 0.001". I'm using the power feed and not touching the compound so the angle of the compound doesn't come into play.

Anyone have and ideas on what could be causing the problem and how to check for it? I don't have much experience with the lathe so I'm kind of at a loss.

Thanks!

It doesn't sound like deflection to me. It is smaller close to the chuck than out at the end. Wouldn't it be the other way around if it was deflection?

I would chuck of a piece of drill rod and mount a dial indicator on the tool post and run it along the length by moving the carriage to see if the headstock is aligned to the ways (for a start).

Alan

It doesn't sound like deflection to me. It is smaller close to the chuck than out at the end. Wouldn't it be the other way around if it was deflection?....
Alan

No, the end deflects away from the tool; in other words the centerline is further away from the cutter so the size is larger further away from the chuck.

Geof,

OK, I understand now. I just wasn't thinking clearly.

Thanks,
Alan

If the lathe has a gap bed make sure the gap peice is seated properly. There may be swarf or a bur on the casting which is preventing the bed from sitting correcly in position, as the saddle of the lathe moves across the section it may be moving away from a level motion resulting in the a taper from the tool rising or lowering as it travels.

Cheers
Splint

Thanks Geof. I think I understand what the deflection is now. I think I will try turning between two centers, but first I need to try to align the tailstock. I will do some reading on how to align the tailstock and give it a go.

If the lathe has a gap bed make sure the gap peice is seated properly. Nope, the 7x10 isn't a gap bed lathe so no problem there. Haha, but after reading about gap bed lathes to find out what they were, I've got crazy ideas in my head about bandsawing parts of the 7x10 off to make a gap bed lathe.

.....I've got crazy ideas in my head about bandsawing parts of the 7x10 off to make a gap bed lathe.

I hope you are joking. No insult intended but this does qualify as a crazy idea.

When you cut significant chunks out of something that is cast you are going to relieve stresses in the metal. There is a chance you would finish up with the ways and the surface the head is bolted to no longer co-planar or parallel.

Gap bed machines have the bed reinforced around the gap and as far as I know the final finishing on the critical surfaces is done with the gap installed so that everything is true.

First try a spring back pass. If you are using a live center in the tailstock, it may be the tailstock is misaligned. You can adjust the tailstock by adjusting the set screws you will find down near the base. Most tailstocks are mounted on a crossways dove tail and the set screws jack it from side to side. just put a piece of stock in and take small cuts while adjusting the tailstock back and forth untill the taper is gone. It's not rocket science!

Your workpiece is extended out of the chuck 4 times your starting diameter and that ratio is growing larger with each turning pass on your way to the desired 12mm diameter (about 15/32"). Use your follow rest or a dead center in the the tailstock if you need accuracy to within .001 of an inch.
Jim

Good God Man ! it's a lathe ! did you go th high school ?

I attended three high schools in three large cities at opposite ends of the U.S. and not one of them had a lathe. Metalworking is not a common course in public schools. Typing is, though.

Do any of your posts have any useable information? Do you know the solution to this guys taper issue? Do you have anything of any merit to contribute? Seriously?

Please find another source of entertainment.

Scott

Condescending remarks seem to be the norm in this thread and becoming common place on the zone in general! Its a shame really.

Condescending remarks seem to be the norm in this thread and becoming common place on the zone in general! Its a shame really.

That is an exaggeration. There is one post in this thread that is silly which are the ones that consitute condescension? My take on most of the posts here is that they either ask legitimate questions or propose legitimate answers.

(response to oldman's post)...and they should be trounced immediately. This forum is for enthusiastic intelligent people - not those who are stuck in puberty trying to figure out what they can get away with from behind their keyboard.

I usually ignore these folks - just like others typically do, but sometimes I just can't.

Sorry to detract from the thread, guys. Let's move on.

Scott

What is really going on!!
Is 1) the runout on the chuck bearings. or 2) the bed could be a little warped...

A live center is a MUST, your TAIL stock could be OFFSET too... I am assuming that it is not with the above information number 1 and 2..

Hackman; To put it mildly you are being given conflicting opinions. To determine whether it is workpiece deflection or machine inaccuracy do an experiment: Get a piece of 2" diameter 6061 and turn off a similar amount. i.e., take off about 0.03". The 2" diameter will definitely not deflect measurably over a length of 2 inches and if you get taper this time then you know it is due to the machine.

Geof
I am not disagreeing with you that it can or can not happen, I am just saying that 1/2" material with only 2" of cut. Seems unlikely to happen with a live center in it's proper place. It takes lots of pressure to bend this short of material.

What is really going on!!
Is 1) the runout on the chuck bearings. or 2) the bed could be a little warped...

A live center is a MUST, your TAIL stock could be OFFSET too... I am assuming that it is not with the above information number 1 and 2..
He was never using a tailstock from what I gathered,

If your bearings are not tight the live center will only be as good as those, either in the live center and/or the spindle. I was taught that dead centers with a trued cut(preferably in a collet) on an aligned center cut are the only true way to get a accurate cut as live centers can have walk in them. This is of course assuming that your spindle bearings are tight. Also if its mounted in the 3 jaw you could easily have a small amount of taper from the jaws and making the assumption that the piece is true itself. The next best thing is the follower rest in this case to help with deflection since its so short.
Also he stated he was not touching the compound, if the slop is not taken up in the compound some undesired effects may come about.
chris

The only other thing I would add is; does the geometry of the cutting tool suit the material type?

John

Hi,
check that the chuck is not loose on the drive spindle.
check that the two beariings in the main spindle are not worn.
Check that the compund slide has no movement.
make sure the tool post is bloted down firmly.

regards

Colin
Toolmaker

It sounds like you're not using a live or dead centre but you should be. Assuming you are and the taper issue is still there, then maybe there is wear in the base of the tail stock or in the tail stock spindle which is causing the live/dead centre to push the end of the job downward. If that is the case you may be able to shim the tail stock up to the point that the centre is where it is meant to be.

Cheers
Splint

There are some great suggestions here, but for my 2 bob's worth, do the simple checks first. If it passes, great, if not move onto checking bearings etc.

The aim of fault finding is to diagnose the problem quickly. If takes a few minutes to chuck up some stock, take a light cut and measure it. If this confirms the fault is in the bearings or alignment, you've wasted (not really wasted as it confirms fault) a few minutes as opposed to checking bearings and alignment which will take considerably longer.

I'm not condemning any of the suggestions above, all are very valid. I am suggesting doing them in the right order to make the best use of your time.

What'e the bob's worth? The Fault finding method will not tell him anything except he has run out on the part...
Plus he has done this repeatedly?

He does need a center if he is not useing one. A stead rest could be setup and used but it will not help determin a problem...other than the machine is twisted or un-level..

Chuck runout could be checked by useing a dial indicator and your hands.. Put the dial 180 deg from where you will be pushing with your hands (or a safe non destructive way). Push it and see if the dial moves..

On a worse note, the machine may not be capiable of attaining better than +/- 0.001. Which is pretty close for a hobbiest.

For me it sounds like your headstock is tilted one way or the other. I have the very same problem on my mini lathe. Mine is only 0.034 degrees, but that leads to 0.06mm per 100mm, which is far to much for me

Hi everyone,

Wow! Thanks for all the suggestions. I haven't had a chance to try all of the suggestions yet, so far I have been trying to align the tailstock so I can try turning between centers. In the original post, I was not using the tailstock at all so that may have contributed to the problem. In the process of aligning the tailstock, I started checking the alignment of the headstock and it looks like my spindle is not aligned properly either. I measured 0.003" horizontal misalignment over 6" so I think a big part of the taper is caused by both the deflection and the misalignment.

Oh yeah, the gap bed idea was a joke. I'm not crazy enough to start cutting parts of the lathe off... yet

I'm going to try to true up the headstock and the tailstock and see if cutting with a dead center makes less of a taper. If that doesn't help, then I'm going to try to turn between centers. And if that doesn't work, then I'm just going to say that my lathe was made on a friday afternoon and move on.

When I order a new lathe, I'm going to write a letter:

Dear Sirs;
I know you can do a fine job of grinding the lathe bed, making the spindle, and gears and screws and stuff, but could you please not bother mounting the headstock on the bed? I know you guys just cannot seem to ever do that correctly because everyone I know has to fiddle with their headstock as soon as they get their lathe unpacked. I thought I would save some time by just doing it myself, and not having to pull those nasty dowel pins out and try to ream them so the headstock is on right. Thank you very much."

;)

That's tongue in cheek :D

Headstock misalignment is synonymous with a twisted bed, which is due to improper adjustment of the levelling of the lathe. In other words, if the user does not replicate the position that the lathe bed was in, when it was on the grinder, it is impossible to determine whether the headstock is misaligned or the bed is twisted out of level.

No lathe bed is immovable. You can bolt down 3 legs, slip a pry bar under the 4th, and flex it. Easily. So why not take advantage of that flexibility to 'put your headstock into alignment' instead of fiddling with the fastening of the headstock?

To set up a lathe, a relatively massive workpiece is required to eliminate flex of the work as a source of error. I prefer to set up a lathe by boring a hole, say 3" diameter, 4" deep. It should measure exactly the same from one end to the other. While its easy to 'fudge' an OD with files and sandpaper, its not so easy to fudge a tapered bore into a straight one.

If the hole bores large on the outermost end, then the carriage is approaching the headstock at an angle. The leg near the tailstock end, on the operator's side should be shimmed up to correct the problem. Its a sensitive adjustment. On a lathe 10 feet long, I'll tweak the machine in with 1/16 to 1/8 of a turn of the screw, after using the best level that I could find. I have no trouble with the lathe turning tapers on free ended chucking, and my lathes are 20 years old or more.

So before you alter, ie., screw up, your headstock alignment, check out whether you can accomplish exactly the same thing by simple adjusting of the levelling screws.

..... I started checking the alignment of the headstock and it looks like my spindle is not aligned properly either. I measured 0.003" horizontal misalignment over 6" so I think a big part of the taper is caused by both the deflection and the misalignment....

Horizontal misalignment by that amount will not cause a taper that you can measure.

Thinkg about it, maybe do a sketch and a calculation: If you started the tool on center and then machined 6 inches in either direction the end result would be that the tool is now .003" above or .03" below center. If you do the calculation for the chordal height of the arc subtended by this amount you will find it is infinitesimal.

The OP is using a 7x10 lathe - no dowel pins aligning the head, but plenty of casting sand sitting under it!

My 7x12 was out by ~0.010 horizontally and ~0.005 vertically over 8 inches when I got it. The tail stock was also ~0.010 high. While it was a pain to align, nothing more than a DTI, a short length of ground rod and some metal foil (aluminum works) is required.

A link to a good method to test and align the headstock.
http://www.gadgetbuilder.com/Lathe_Align.html

Aaron


When I order a new lathe, I'm going to write a letter:

Dear Sirs;
I know you can do a fine job of grinding the lathe bed, making the spindle, and gears and screws and stuff, but could you please not bother mounting the headstock on the bed? I know you guys just cannot seem to ever do that correctly because everyone I know has to fiddle with their headstock as soon as they get their lathe unpacked. I thought I would save some time by just doing it myself, and not having to pull those nasty dowel pins out and try to ream them so the headstock is on right. Thank you very much."

;)

That's tongue in cheek :D

Headstock misalignment is synonymous with a twisted bed, which is due to improper adjustment of the levelling of the lathe. In other words, if the user does not replicate the position that the lathe bed was in, when it was on the grinder, it is impossible to determine whether the headstock is misaligned or the bed is twisted out of level.

No lathe bed is immovable. You can bolt down 3 legs, slip a pry bar under the 4th, and flex it. Easily. So why not take advantage of that flexibility to 'put your headstock into alignment' instead of fiddling with the fastening of the headstock?

To set up a lathe, a relatively massive workpiece is required to eliminate flex of the work as a source of error. I prefer to set up a lathe by boring a hole, say 3" diameter, 4" deep. It should measure exactly the same from one end to the other. While its easy to 'fudge' an OD with files and sandpaper, its not so easy to fudge a tapered bore into a straight one.

If the hole bores large on the outermost end, then the carriage is approaching the headstock at an angle. The leg near the tailstock end, on the operator's side should be shimmed up to correct the problem. Its a sensitive adjustment. On a lathe 10 feet long, I'll tweak the machine in with 1/16 to 1/8 of a turn of the screw, after using the best level that I could find. I have no trouble with the lathe turning tapers on free ended chucking, and my lathes are 20 years old or more.

So before you alter, ie., screw up, your headstock alignment, check out whether you can accomplish exactly the same thing by simple adjusting of the levelling screws.

So before you alter, ie., screw up, your headstock alignment, check out whether you can accomplish exactly the same thing by simple adjusting of the levelling screws.

I think that if I had a proper lathe then I would go that route. The cheap chinese 7x10 doesn't have leveling screws, just rubber feet so no dice there. I actually took the headstock off the bed once already to repair some stuff so the headstock is not that bad to fiddle with. I think because the lathe is cheap and has a lot of problems, that I am learning more about lathes then I would have ever learned with a properly setup machine.

A link to a good method to test and align the headstock.
http://www.gadgetbuilder.com/Lathe_Align.html

Been reading that for the last few days and experimenting. Finally gave up trying to use the magnetic base to hold the dial test indicator and milled a proper holder. With the magnetic base I couldn't get repeatable results, but with the holder I can. Also found a nice chromed rod inside a junked scanner to use as a test bar. I attached some pictures.

So after measuring everything, it looks like my headstock is not horribly misaligned. It's canted 0.0027" towards the front at 6" and droops 0.00125" at 6". The droop could be actual headstock misalignment or just the rod sagging. I'm not sure how to tell the difference.

Thinkg about it, maybe do a sketch and a calculation: If you started the tool on center and then machined 6 inches in either direction the end result would be that the tool is now .003" above or .03" below center. If you do the calculation for the chordal height of the arc subtended by this amount you will find it is infinitesimal.

I took your advice and made a sketch. The misalignment really doesn't seem to contribute much to the problem. In fact, I think since the headstock is canted towards the front of the lathe, i.e. into the cutting tool, the far end of the rod should have had the small taper instead of the near end if the misalignment was coming into play since the far end would have been closer to the cutting tool.

At this point I think it probably was the deflection from having the thin rod hanging out so far from the chuck which caused the original taper. I think my next step is to order 2 dead centers, align the tailstock, then try to turn the rod again between centers to see if I can get the accuracy I need.

Thanks everyone for all their advice so far. I'll post the results after the dead centers arrive.

Order 1 LIVE and one dead....
A dead center on the tailstock will cause problems if you have much time on the machine... Heat, Wear, and some instability because you can't really CLAMP it TIGHTLY for a long time for sure...

Check tail stock position if you haven't already.

If you're not using tail stock, perhaps you should as this will cut down on deflection of long slender pieces cantilevered out from the jaw.

A high or low tail stock will induce taper.

A tail stock that is moved "in" or "out" from the chuck C/L can induce taper.

Perhaps a "twisted bed" condition exists which can also lead to some bizarre tolearnce stacks.

Sorry if the above are repeats, I didn't take the time to read the whole thread.

If you want dead nuts accuracy a dead center in the tail stock is the way to go. Spen a few extra bucks for one with a carbide point and use high pressure grease (axle bearing grease) for lubrication.

Wanted to give an update on the taper problem. I picked up a 3MT dead center and a 2MT dead center along with the 2MT live center which I already had, so I think I have all the types of centers covered. Been spending the last few days trying to get the tailstock aligned on the 7x10 and I got it within 0.001".

So I did an experiment where I have the same bar chucked 1" into the 3 jaw chuck and have 2" of the bar exposed. Used the tailstock and started turning. The bar is 1/2" 6061 aluminum which I am trying to turn into a cylinder accurate within 0.0005" in diameter along the 2". The results are... Diameter of the bar near the chuck is correct, diameter of the bar near the tailstock is correct, center of the bar is larger then the ends. Does it sound correct that the ends are anchored correctly now but the center of the bar is deflecting? Would turning between two centers make any difference? I imagine that the center of the bar would still deflect even when turning between two centers.

So this makes me think of new questions. Is it possible that the 6061 aluminum is not the right material for this job since it is too springy? Would something like drill rod deflect less and make the problem go away?

I've already thought of some other ways to make the part that don't need such an accurate rod, but now this is becoming an obsession as to how I could actually turn the aluminum rod into a near perfect cylinder if I had to.

..... The results are... Diameter of the bar near the chuck is correct, diameter of the bar near the tailstock is correct, center of the bar is larger then the ends. Does it sound correct that the ends are anchored correctly now but the center of the bar is deflecting? Would turning between two centers make any difference? I imagine that the center of the bar would still deflect even when turning between two centers....

Yes, if you are turning between two centers you can expect the bar to deflect. You can look at it as a simply supported beam with a load in the middle; it will deflect.

I would expect the deflection to be less but I am not sure: Again look at it as a beam; your first try with the part held at only one end was equivalent to a rigidly held cantilevered beam. You would need to do some sample calculations to see how the deflection at the end of a cantilevered beam compares with the deflection at the center of a simply supported beam under the same point load; at the end or in the center.

There is another potential source of deflection when turning between centers and that is thermal expansion. You crank the center in with the part cold, when it is turned it warms up and expands but the centers are preventing it getting longer so the expansion causes it to bow. This is not likely to be relevant to your situation but there is a slight possibility.

Regarding using a different material you may get better results; but only may. The stiffness of any material depends on Young's Modulus E and E for aluminum is about ten times less than steel so the steel should deflect less. However, the steel is harder and stronger so the cutting forces are greater so the deflection may not be reduced as much as may be expected.

Sometimes it is not a bad idea to be obsessed with something because it can turn into a valuable learning process...just don't become too obsessed. When you start dreaming about turning aluminum cylinders you have gone too far:) .

I am quite frankly amazed that you are getting >.001 with a three jaw chuck setup. They can often be out as much as .003-.005 esp if they are well used or cheap. My machine tool teacher refered to that method as the lazy mans centers, I was trained not to machine, but be able to fix them.

I would at this point get a known good ground piece of steel(say maybe a reamer blank or drill rod of better qaulity), and chuck or mount it between centers of which the latter will be more accurate. Watch your TIR if all is well it should be very small(try it applying some pressure too.), I think the tolerance on the lathe is .001 between centers so you are asking alot with out some modifications and tightening anyway.....

I personally would be prazing god if I got .002 out of my 9x20, but that is a different and far less sturdy beast.....

If I were to take a geuess its one of two things, my first geuess is runout from the spindle bearings(they are good but not great), which will get worse the farther away form the chuck you get, this is normal when mounted in a chuck, as chucks have alot of surface area to transfer the smallest amount of runout(since you are grabbing the side and not the face. When you are between centers the runout is considerably less and tends to self align, as I stated in an earlier post preferable between two dead centers on of which is cut to the spindle(you take a turing face cut down the taper each time you mount it, I prefer 12L12 myself) and lubricate, I use linseed oil and white lead, you may not have the option. Its easy to run it between centers with the spindle end in the chuck, it would be better to run it into the spindle taper(requiring a tapered mount) or a collet but the results should be nil to very small if you true the center in the chuck. Hard centers are nice, but a little less accurate if they happen to be untrue or there is a problem between the spindle and the taper etc.

The other geuss is what you are saying with deflecting material, alum is not ideal for checking, but if thats the material you want to machine you may have to over come the problem anyway, the follwer rest would be my first suggestion, lighter cuts, changing tooling, and use of coolants may also effect this some.
It could also be a warped bed/worn ways, but I find that less likely, you would see that however if you run a ground bar with the test indicator as stated above.

chris

Hi Hackman.

Please try this, having turned your part with the .001" error, replace the lathe tool with the DTI. But do NOT put the stylus on top of the piece put it at 90 degree's "Imagine it as a cutting tool" then run the stylus along the part to see whether the DTI is following the exact path of the part, if it is you now know it isn't the part deflecting due to the cutting action as you are not cutting.

The other thing to check with the same setup if it is possible to use the power feed as the leadscrew be out of line pushing up on the saddle, you would have to have the motor/chuck not running if it is possible with your machine.

HTH
Phil_H

Hi Hackman.

Please try this, having turned your part with the .001" error, replace the lathe tool with the DTI. But do NOT put the stylus on top of the piece put it at 90 degree's "Imagine it as a cutting tool" then run the stylus along the part to see whether the DTI is following the exact path of the part, if it is you now know it isn't the part deflecting due to the cutting action as you are not cutting......

Now why didn't I think of that??? (I need an emoticon for "chagrined")

With only 2 inches of 1/2" OD stock sticking out of the chuck and a good measurement at both chuck and tailstock ends suggests a deflection caused by the tool tip. Can't see it being a machine setup fault in such a short distance unless the bed is grossly twisted or worn.
To cut accurate on these little machines you need sharp tools and light DOC with 1/2" aluminum. The tool must cut freely and not rub. A good very sharp hand honed tool WILL cut less than .001" DOC.
I use a diamond hone on solid carbide boring bars to take off a nice curly chip that can only be seen with a magnifying glass for instance when I sneak up on a bearing bore.
Also you will want to learn to use a file to true up parts from deflection as many times it is simply faster and easier. There is nothing wrong with a little hand filing in the end.
You can measure as Phil_H advises to rule out any gross machine error for peace of mind.

Steve

Hackman,

As a general rule of thumb, you can only let your material stick out of the chuck 3 times the diameter. Now, there are exceptions to that rule, but it all depends on the material type, and just what you're trying to do.

You can probably get away with cutting the 12mm diameter, because it is aluminum, and you're only removing .02756". However, you will need to have an extremely sharp tool. A tool with any kind of radius, or honed edge won't work. You have to remember that metal of any kind, including brass and aluminum, doesn't particularly like to be cut. It *will* try to move away, and a radius or honed edge will only assist it in it's attempt to move.

Since you seem to need to have this cut straight (didn't like the .001 taper you had), you will probably need to take a skim cut of .001 or .002. For that you'll need razor sharp cutting tool. A piece of HS, sharpened on a green wheel won't do it.

Unless, of course, you cut it between centers. Then you can cut it in one pass with most any tool that has a decent edge.

Matt

Small lathes have a lot of flex built in. I can see the effect of a small amount of pressure in a number of places on my G4000, not the least of which was the compound slide assembly till I converted to CNC and just junked that part as more grief than it was worth.

Rig your dial indicator again and observe the effect of pressure and torsion against diverse parts. It will be an eye opening exercise! ; )

So I did an experiment where I have the same bar chucked 1" into the 3 jaw chuck and have 2" of the bar exposed. Used the tailstock and started turning. The bar is 1/2" 6061 aluminum which I am trying to turn into a cylinder accurate within 0.0005" in diameter along the 2". The results are... Diameter of the bar near the chuck is correct, diameter of the bar near the tailstock is correct, center of the bar is larger then the ends. Does it sound correct that the ends are anchored correctly now but the center of the bar is deflecting?

It sounds like he used tailstock support and had a larger measured diameter in between the 2 supported ends on a part of only 2" long.
IMHO this is caused by a dull poorly cutting tool.

Yes lot's of flex in the small machines. Pessure from just one finger in the right area can have a profound effect. You can't be an ape with these little machines.
Steve

Just the force of the tool starting the cut can cause some deflection. And then,toward the end of the cut, tool pressure brings the tool back onto dimension.

Dull tool, cutting edge too high or low, improper angle of approach of the cutting edge.... Any of these things COULD cause the symptom.

When I want something super tight tolerance, I stop short of the desired dimension. Then make several passes with no movement of the tool, to let tool pressure cut to zero. If the part is still oversize, I finish with a file and/or emery cloth.

But then those are just my redneck methods!! = )